NEDO Greater Manchester Smart Communities Project Final Report Hitachi, Ltd. Daikin Industries, Ltd. Mizuho Bank, Ltd. 21 st November 2017 This experiment is subsidized by the Ministry of Economy, Trade, and Industry, and by the independent administrative agency NEDO (New Energy and Industrial Technology Development Organization).
1. INTRODUCTION
Background of the Project United Kingdom - Promote Energy shift from Gas to Electricity in the Residential Heating Market Along with the pervasion of High Efficiency Heat Pumps, brought in Demand Response Systems for - Conserving Electricity, Power Grid and - Arbitration for Power Balancing function Heat Pump is Demand Response is Heat Outside = 2 Electricity Input= 1 Heating/DHW Output = 3 Demand response is a change in the power consumption of an electric utility customer to better match the demand for power with the supply. From Wikipedia 3 /44
Demand Response Structure Heat Pump Aggregation System Residential Heat Pump Capacity Shortage... Saving Request Hitachi Electric Power Aggregator Nega-Watt Request DAIKIN Heat Pomp Aggregator Aggregate Small Nega-Watts Widely Utility or ISO Unstable Planning Modification Execution Report DR Communication using OpenADR2.0b Energy Consumption Data as DR evidence DR Trading @ Market Contract Fee Incentive 4 /44
Location Maps 5 /44
2. INSTALLATION OF HEAT PUMP UNITS & THEIR DEMAND RESPONSE AGGREGATION 6 /44
Gus Hybrid Monobloc Split Heat Pump Systems : Simplified Diagram Normal + Buffer Vessel N/A 7 /44
Installation Result Total of 550 Heat Pump Systems installed in Greater Manchester Area Normal Units of Electric Heat Pump, LT Split/Monobloc, and Gas Hybrid And Heat Pumps with Buffer Tank for Heating Among the Electric Heat Pump Split Units, 10 Units are from Hitachi Normal Units With Buffer Tank LT Electricity LT Hitachi Gas Hybrid Electricity Gas Hybrid Mono Hybrid LT Mono Hybrid Sub Total WALH 161 8 127 7 3 1 0 307 NWH 70 2 27 35 19 0 0 153 STH 0 0 15 75 0 0 0 90 Sub Total 231 10 169 117 22 1 0 550 Category Total 410 117 23 0 550 8 /44
Site Surveys, Installation Jobs, and Units 9 /44
Scenery: Installation Completed Sheltered Flats at Wigan Council Daikin Engineer checks settings Sheltered Flats at Northwards Housing 10 /44
Demand Response Trials Period: October 2015 to March 2017 DR command: Turn off the heating units DR schedules: Twice a day, morning and evening Duration DR server side 1 hour to 2 hours Property side: 30 minutes to 2 hours 11 /44
Demand Response Result: 27/01/17 Demand Response reduced around 50% of Electricity than the assumed Base Line. Total Power Profile over 460 properties, 27/01/2017 Date of Experiment 27/01/2017, Friday, AM7:00~AM9:00 Outdoor Temp -2.9 Nega-Watt by DR 234.5 kw 12 /44
HP Heating System for the DR Resource Each Heat Pump Unit starts/stops intermittently, but once these units are aggregated, the whole system provides a good secure NEGA-WATT value. Within two minutes of the DR event kicking in, it reduces to the target amount. Responsivity to the DR commands seems satisfactory. Heat Pump System can be valid to Fast DR resources. Energy Total Sum HP Total HP1 HP2 Each Unit Energy Usage HP3 HP4 13 /44
Safety Functions for Tenant Comfortability Any Property connected to DR server is DR Application Property, for the Project During the DR event, it stops Heat Pumps running, and Heating Capacity decreases. To keep Tenants comfortability, Safety Functions are built in the DR Program. Safety Function; Under any of the following conditions, the property is excluded from the DR event: When the Room Temp is/becomes below 18, at any time of the DR event. When the Room Temp drops 2 from the beginning temperature. When the tenant(s) touched and changed the Set Point at the Remote Controller. Safety Function for Room Comfortability ( Automatic Opt-Out ) Opt-Out ( Manual Opt-Out ) Room Temp < 18 d. C Room Temp drops 2 d. C Tenant(s) changes the Set Point When the DR Event Begins Excluded from the DR If the Temp > 18 d. C, DR event starts If the Temp > 18 d. C, DR event starts During the DR Event Excluded from the DR When the Temp drops 2 d. C, excluded from the DR No matter which temps, tenant(s) want to change the Set Point, DR ends as Opt-Out. 14 /44
Numbers of Break-Away from DR Break Away from DR : Around 1/3, Totally 27/01/2017 7:00AM-9:00AM The DR system is designed on Opt-Out basis. All the properties participate in DR. When Tenants change the set point at the Remote controller, it is recognized as Opt-Out. This system secured the Electricity Reduction Management. Break Away from DR were 151: From 364 to 213. And 119/151 were Safety Stops. Numbers of Break-Away from DR and the Causes 27/01/2017 7:00AM 9:00AM Total Participants: 364 Safety Stops: 119 Opt-Out: 32 Time Scales after the DR Event start (Minutes) Safety Stops Opt-Out by Users request 15 /44
Internet Disconnection at Properties High Speed Internet, ADSL wired the Heating/Monitoring Units to DR Servers via Broad Band Routers, but tenants frequently pulled out plugs or turned them off. Those communications are beyond Heat Pump Service Engineers, and internet engineers visited over 300 times to re-connect cables and re-boot. Mobile devices for Residential IoT units are highly recommended. Visit to Properties to re-plug cables. More than 300 times, but still decreasing. 16 /44
3. FINDINGS ON HP POWER CONSUMPTION AND DR RESULTS 17 /44
3. Findings on HP consumption and DR results 3-1 HP power consumption 3-2 Demand response results 3-3 Tenant acceptance 18 /44 Hitachi, Ltd. 2017. All rights reserved.
1. HP power consumption Effect of aggregation Individual Profile (9th Nov. 2016) Aggregated Profile (9th Nov. 2016) Peaks Num. of tenants: 372 Individual profiles are quite different Daily Habit Set point of internal temperature Intermitted Energy saving by HP local controller Aggregated profile shows a pattern two peaks in morning and evening HP power consumption uplifts total Hitachi, Ltd. 2017. All rights reserved. 19 /44
1. HP power consumption External temperature is a major factor All HP type combined Electric HP Gas Hybrid HP External temperature is a major factor of HP power consumption Hitachi, Ltd. 2017. All rights reserved. 20 /44
1. HP power consumption Developed an estimation model with mix of HP types Electric HP (Actual data) Electric HP model Gas Hybrid HP (Actual data) Gas Hybrid HP model Developed an estimation model for HP power consumption by each type External temperature HP power size ( e.g. 4kw, 6kW, 8kW ) Number of HPs Actual value vs Estimated value (550 HPs) Actual HP Power Consumption Estimated HP Power Consumption Strong relation was observed between estimated values and actual values X axis: Estimated values by the model supposing 550 tenants Y axis: Actual HP power values adjusted to 550 tenants Hitachi, Ltd. 2017. All rights reserved. 21 /44
2. Demand Response Results Three parameters of DR Examples of DR Profile (Aggregated) DR Num. of target housing Period Total DR trial times DR Trials 4 550 housing 18 month (Oct. 2015 Mar.2017) ~ 360 DR Parameters Power [W] Base Line HP Power Response time DR amount DR duration Time Hitachi, Ltd. 2017. All rights reserved. 22 /44
2. Demand Response Results Response time DR amount in 1 st SP Definition of Response time Histogram of Response time Power [W] Base Line 1 st 30 min 2 nd 30 min 3 rd 30 min 4 th 30 min HP Power Mean pow in 1 st SP Response time Time Response time is; Time between DR start time and time HP power first reaches the average of the first settlement period Median Average Standard deviation 2 min 2.3 min 0.75 Maximum 6 min Hitachi, Ltd. 2017. All rights reserved. 23 /44
2. Demand Response Results DR Duration DR amount DR amount DR amount DR amount Power [W] Base Line Definition of DR amount Commercial Rule of Flexitiricity (UK aggregator) An example of 90 min duration DR (illustrative) 1 st 30 min 2 nd 30 min 3 rd 30 min 4 th 30 min HP Power [%] Success rate of DR duration 95% of 1 st 30 min DR Success Failure Time [min] DR duration If DR amounts of following settlement periods keep under 95% of the first one, DR successfully continues Continuous (auto) Opt-Out decreases DR amount In the condition of the project, 60 min is relatively stable duration time Hitachi, Ltd. 2017. All rights reserved. 24 /44
2. Demand Response Results DR amount Histogram of DR amount Actual value vs Estimated value (550 HPs) Actual DR amount Estimated DR amount Target DR amount (200kW) in this project was achieved 144 times in actual values Hitachi, Ltd. 2017. All rights reserved. 25 /44
DR amount [ kw] 2. Demand Response Results DR amount estimation through a year Monthly HP sum power for 3hrs (6am-9am) Adjusted to 550HPs, Oct.2015-Mar.2017 4-6 times Big variance of HP power through a year Estimated DR amount Monthly variation of HP power consumption Adjusted to 550HPs, Apr.2016-Mar.2017 DR amount is calculated with the estimation model using actual external temperature 90kW 341kW Maximum and minimum DR amount among monthly average ; Jan. 341kW (621W/tenant) Aug. 90kW (164W/tenant) Hitachi, Ltd. 2017. All rights reserved. 26 /44
Power 2. Demand Response Results "Reactive Operation" of DR Definition of Reactive Operation B A Reactive Operation of DR is a phenomenon where B is higher than A A : Pre 2 min mean of DR ( ~ baseline) B : Post 30 min mean of DR Histogram of gap between A and B A < B Reactive Operation happened in almost all case of DR with HP Mechanism of Reactive Operation Room temperature Set point 1 2 DR Temperature Time 1. 1 Temperature goes down during DR Note: Drop of temperature is less than 2 degree (C) by Safety Function 2. 2 Local controller turns up HP immediately to recover internal temperature back to the set point Hitachi, Ltd. 2017. All rights reserved. 27 /44
3. Tenant Acceptance Automatic Opt-Out (Safety function) vs Manual Opt-Out Opt-Out ratio External Temp.(green) Total Opt-Out (Auto + Manual) ratio Ave. 10.6% Standard deviation 7.3% Manual Opt-Out Ave. 5.3% Standard deviation 4.3% Manual Opt-Out (blue) DR Event ID Automatic Opt-Out (red) Correlation coefficient Ext. Temp. vs Auto O/O -0.71 (Strong relation) Ext. Temp. vs Manual O/O -0.39 (Weak relation) Hitachi, Ltd. 2017. All rights reserved. 28 /44
3. Tenant Acceptance DR awareness DR awareness Ratio of tenants who didn t notice DR operation (%) ALMO Never, Blank, Rarely All 89% Northwards 87% SixTown 93% Wigan&Leigh 87% Sample size 70 Hitachi, Ltd. 2017. All rights reserved. 29 /44
4. BUSINESS MODEL ANALYSIS 30 /44
Business Model Analysis New Co. is a company to manage DR of HP mainly for public facilities and social housings. New Co. trades nega-watt aggregated from HPs as Balancing service. The economic evaluation of the New Co. is calculated based on measured data in the project and other assumptions. New Co. nega-watt trading Balancing service TSO mainly STOR market DR Control Local Council Local Council Local Power Retailer own own Public Facilities Social Housing HP Step1 (2018~) Aggregation of assets in public facilities in GM Aggregation of HPs in social housings in GM Social Private Housing Subscriber HP HP Step2 (2022~) Aggregation of HPs in private houses Aggregation of HPs in social housings in areas out of GM 31 /44
Business Model Analysis To cover energy management system cost and working capital with income from Nega-watt trading, 55-60 thousands HPs are necessary to participate, which will take long term to realise. While DR based on HPs is a potential field, the business model that relies solely on DR based on HPs are limited. It is necessary to combine with other businesses such as ESCO business, or incorporate other DR sources to stabilize the business. 5.0 4.0 3.0 (mil. GBP) revenue : public facilities revenue : HPs in social housings revenue : HPs in private houses revenue : HPs in areas out of the GM operating income 2.0 1.0 0.0-1.0-2.0 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 32 /44
5. FUTURE WORK 33 /44
Future work Post NEDO project started from Oct. 17 Data accumulation of HP power consumption for further analysis DERs(Distributed Energy Resources ) will increase more in UK (e.g. EV, PV) Integrated management with HP and EV could be an interesting solution for; Business model Optimisation of network reinforcement Hitachi, Ltd. 2017. All rights reserved. 34 /44